Finalist | Excellence in Narrative

Stephen Alexander and Jordan Thomas are two of the creators of The Magic Circle.

Each year at GDC, MFA students from the NYU Game Center interview the Independent Games Festival nominees, asking them three questions about their development process. In addition to this interview, you can read all the insightful interviews from 2016 here. These conversations, and much more, will happen when the Game Center returns to GDC in 2017. Learn more about the Game Center at GDC 2017.

Alexander King: All game projects start with some kind of spark, from a design question to a feeling you wanted to evoke. What was the spark that grew into your game?

Jordan: Well the initial spark was irreverent white board graffiti at game developer’s, right? So people kind of sniping at each other with usually way too elaborate, way higher production values than they have any right to be doodles. Subversive doodles… at various companies we’ve worked for. I guess there’s a slight caveat to that which is that much earlier than that, while working on Thief III actually, way back when, I remember seeing all these guys standing around in T-pose, right, and realizing that I didn’t know what I was doing and trying to finish my level and I… felt bad for them. And I imagined them looking up and saying, “C’mon buddy! Are you gonna finish this year?” sort of stuff. That was just lodged in my brain. And then we finished Bioshock Infinite, and had done lots of BioShocks essentially, and our AI programmer had done Dishonored, we felt like it was time to do something sillier, or at least, lighter of spirit. So we discussed a few different pitches that were very very dark, and discarded those in favor of this notion that maybe game developers could be arguing pointlessly over a world in progress, and you the hero would wake up, and like, essentially the game would wake up and ship itself. There’s something kind of fun about that.

Alexander: On a lengthy game projects, many developers say they enter what you could call a “valley of despair”. Did you experience this during your development process and how did you push through it?

Stephen: That should’ve been the name of a location! I think that in the sense that, well yes, but there wasn’t much time for that. I think there was definitely a point where we were like, okay we just have to get this done. And there was that period of sort of frustration with stuff we were currently working on, and sort of banging your head against it, stuff like that. But, we also, being indie, and irresponsible, would continually give ourselves room to keep adding new ideas even when we probably shouldn’t have been, which at least for me staved some of that off. Because it wasn’t just like, this is what the thing is, let’s just churn. We would have discussions, especially in response to periodic “Hey friends, here’s where we’re at, let us know” to reexamine and potentially change in some ways. There were definitely micro stretches of what you’re describing. I still remember getting into, “Now, I just have to make every permutation, of every AI, for every ability…” And the thought of it was actually more of a valley of despair than the actual doing of it. Standing there in front of it, thinking, my next month is just this. Then I would start to find joy in the actual individual problems. Jordan could probably speak more to that, I know he on script I think ran into that more than I did.

Jordan: Oh yeah. It’s actually a… deeply unflattering story because in this one case my doubts turned out to be well founded. Throughout, I mean we decided to write a game about games, and we ran into a lot of different semantic issues. You know like, who does the developer refer to? Us? Or the fictional developer? What does ‘the game’ mean? All of this stuff that only happens if you get this meta. And there were various moments where I was like “Maybe this was just a really bad idea” because the amount of decoding that a person has to do to enjoy it is quite high. Now, for its people, it’s like crack. You watch them just grin ear to ear. But we have found that, with the sales on Steam and so forth, it is a heady pitch. And people kind of need to be sold it by a friend, and word of mouth’s really hard now because there’s just so damn many games on Steam. So, we’re very proud of it, but on the other hand, those doubts about how big the audience was were pretty much accurate. We are moving to Playstation and Xbox One soon in the hopes of it finding its people on all platforms.

Alexander: And lastly, how different did that end product end up being from your original vision?

Jordan: Well, Stephen heroically figured out a way to end it. At some point, this pitch trailed off into… “and ideally, ‘spoiler thing’ would happen” but there was no way we could afford to do that. Near the end of the game, he was like, “So last night I sort of put that thing in”, and I guess all I can say is that it delivers on the fantasy that the game implies in a way we thought was impossible for a team of our size. And heroic work between Stephen and KANE CHIEN?, our AI programmer, it makes the game so much more complete and so much more satisfying. And I did not dare to conceive that we could get that far with it when we were talking about the original plan.

Stephen: There was definitely no concrete representation of… it was sort of a spiritual, “this is how it should maybe work” but almost immediately, I don’t even remember it being articulated, we backed off of it, and were talking about how can we end this thing as escalation. And both of us knowing as we discussed these ideas that they all felt not satisfying or right. And then it turned into this other thing…

Jordan: I know we’re talking around it, but it’s one of the big discovery fun moments of the game, so we don’t want to spoil it.

Alexander: Circling back to something you said earlier, you talked about avoiding the valley of despair by being able to add features through development. That’s an unusual response, people talk about needing to cut down whole parts of the game…

Stephen: Usually, that’s what people do, and especially if your goal is extreme polish then you kind of have to do that. With our game, we early on owned the notion that there would be rough edges. The key was that the player’s experience couldn’t have a ton of tough edges. The presentation could be a little bit rough because it’s on theme. As long as it was artfully rough.

Jordan: And a good example of that, is early on Stephen animated a fun scene where one of the game developers is sort of using an in-game model as a giant puppet. And it’s crudely animated because he doesn’t have any animation to use. And people tend to laugh at it anyway, they don’t mind, because they could tell that the character has limited resources.

Stephen: And on that note, that particular instance is one where I could’ve done exactly what an actual developer would’ve done and just robot around. But in order to make it at least more interesting on the art side I actually gave him little bits of doll flop, to make it that that much more apparent that this guy is playing with dolls. That’s what developers often do, you know, they do the Space Balls “Lord Helmet” scene. And so I tried to find ways to just inject a little bit of what you might call polish so that the experience would feel more complete, even if we knew that we couldn’t achieve everything that was in our minds and deliver something that was as polished as something like Mini-Metro or SuperHot or things like that. We would think, “Oh this will make the player’s overall experience better if we put in this new ability” which opens up a whole new path of combinatorial solving. And any responsible team would’ve said, “No. You’re now in full production, the time for those ideas is gone. You could’ve had that six or seven months ago, and it would’ve displaced something else.” But crucially, and probably changing the future, we didn’t really have anybody who was a production mind. And so when the three of us got excited about something, that was sort of enough for us to push it through.

Jordan: Part of it is cultural, or philosophical. We have this, on the one hand it’s the thing that differentiates us. On the other hand, you can almost never completely control scope. We come from this genre of games called ‘immersive sims’, we really just means usually first-person games with lots of tools, lots of ways to solve the problem, which is usually… ending the life of another being [Laughs]. We wanted to do that, but through a creative filter in this game. So all the puzzles in the game have many right answers, we want you to feel like the creative one instead of the designers looking good. But the way to do that is to multiply the tools that you have to bring to bear on any given situation. So there was a certain problem with scope control, just making you feel free enough to pick your favorite tool.


Alexander King was once an analytics and strategy consultant, who used Excel, statistics and common sense in order to improve businesses. Now he puts those skills to much better use in making games!